Shuttle program extension?

How long would it take to build one or two new shuttles from scratch ?
Same core design, but with some of the improvements wanted by NASA (such
as electric APUs).

Someone had mentioned that it would cost about 2 billion to build one.
(perhaps another billion for the second shuttle ?).

This would allow NASA to implement many of the improvements to reduce
costs, and retire the older shuttles instead of having them go through
the recertification and major maintenance cycles needed after they've
done ISS assembly.

Would building new shuttles end up costing same ballpark as rebuilding
the current ones ?

They could keep one shuttle pad and maintain a few shuttle missions to
LEO per year (to ISS' hubble etc).

They could develop a re-entry capsule to be used as ISS espace pods
(brought up by shuttles), and later scale those capsules up to be able
to go to the moon on some new rocket.


If NASA were to gear down to support only 2 or 3 shuttle flights per
year, could it seriously lower its fixed costs on the ground ? I am
thinking that if fast turnaround were no longer needed, wouldn't they
require far fewer workers ? And with only 2 shuttles, wouldn't that free
up some buildings used for shuttle maintenance ?